RAPID RESPONSES FROM BMJ.COMBMJ 2003; 327 doi: https://doi.org/10.1136/bmjusa.03020003 (Published 19 November 2003) Cite this as: BMJ 2003;327:E192
From BMJ USA 2003;Feb:82
Following are edited excerpts from Rapid Responses generated by this article, which can be read in their entirety at http://bmj.com/cgi/eletters/325/7377/1387.—Editor
Headline not justified
- C S Ripley, radiologist ()
EDITOR—This study does not justify the headline “Early CT improves diagnosis in abdominal pain” [This week in the BMJ. BMJ, December 14, 2002] when there was no statistically significant difference at 24 hours between the two groups in terms of accuracy of diagnoses. Furthermore, the length of stay was not reduced by a statistically significant degree and there was no statistically significant change in the number of operations.
It is claimed “Early computed tomography did, however, identify significantly more of the serious diagnoses than standard practice, and it was probably this aspect that affected mortality.” However, data reported for the standard practice group suggest that the mortality from surgical causes was related to perforated viscus and ruptured abdominal aortic aneurysms. There were no ruptured abdominal aortic aneurysms detected by early CT and only one perforated viscus, which was treated conservatively. Put another way, CT was not shown to have an effect on mortality.
Early abdominal computed tomography
- Anusha G Edwards, research registrar,
- Andrew R Weale ()
EDITOR—The authors have not given the grade of surgeon or radiologist involved in patient assessment and reporting of the scans, respectively. An alternative explanation for the 1.1 day shorter stay is that the study was conducted during an off hours period to ensure access to CT within 24 hours. However, the standard practice group did not have any of their investigations expedited as part of the study. Therefore, the apparent increased length of stay in the standard practice group may be due to longer waiting times for investigations and their subsequent reporting.
The difference in mortality between the two groups may perhaps be explained by the difference in case mix. The standard practice group included five patients with a perforated viscus and one with a ruptured abdominal aortic aneurysm. These patients would not necessarily have had improved survival as a result of early CT. This study, as the authors point out, was not designed to investigate mortality, and such chance findings are perhaps to be expected with this sample size.
Early CT scan: yes, but after the evaluation by an experienced clinician
- Paul J Willemsen, consultant surgeon ()
EDITOR—Only patients who did not need emergency surgery or other urgent interventions were included. The mortality in the standard treatment group might be explained by the fact that patients in this study were recruited during weekend hours and that these patients are seen by less experienced staff.
- Peter F Jones, emeritus clinical professor of surgery ()
EDITOR—At least one-third of patients with acute abdominal pain have a self-limiting condition, known as acute non-specific abdominal pain, which settles without treatment in 24–48 hours. It is a questionable practice to submit all these patients, mostly young, to a substantial dose of ionizing radiation before allowing time for observation.
Ng et al correctly advise that “computed tomography should … be used with caution” and be reserved for difficult cases. In a prospective series of 625 comparable patients in New York, Weyant et al found that “there was no correlation between computed tomography findings and pathologically proved disease”; they advised “more precise patient selection by clinical criteria” (Surgery 2000;128:145).
Before extending the use of early computed tomography, consideration should be given to the use of a disciplined regime of clinical observation (with blood tests and plain radiography as required). In over 2000 patients this regime was safe (one death) and allowed patients with a perforated viscus or other active surgical condition to receive timely recognition and operative treatment (Br J Surg 2001;88:1570).